SMART INDENTATION METHODS: THE APPLICATION OF NEURAL NETWORKS |
|---|
| N. Huber, E. Tioulioukovsski |
- Abstract:
- In the last decade, the nanoindentation technique has become one of the most important characterization methods in micro dimensions. The experimental and analytical techniques have been pushed towards an identification method that can compete with tensile tests. It is self-evident to apply these powerful tools in macro dimensions as well, where the nanoindentation technique has its roots. In this paper a new method is presented how the true stress-strain curve as well as the viscosity and creep behaviour of a given material can be extracted from the indentation curve by using a smart analysis tool based on neural networks. Finite Element simulations are carried out for randomly chosen sets of material parameters and maximum indentation depth. The resulting load-depth and depth-time curves are collected in a database together with the material parameters. With this database neural networks are trained to identify the material parameters from measured load-depth and depth-time curves.
- Download:
- IMEKO-TC5-2002-007.pdf
- DOI:
- -
- Event details
- IMEKO TC:
- TC5
- Event name:
- Joint International Conference on Force, Mass, Torque, Hardness and Civil Engineering Metrology in the age of globalization
- Title:
- 8th HARDMEKO Conference on Hardness Measurement (together with 18th TC3 Conference on Force, Mass and Torque and 1st TC20 Conference on Civil Engineering Metrology)
- Place:
- Celle, GERMANY
- Time:
- 24 September 2002 - 26 September 2002